A flux gate sensor =93modulates=94 the ambient field.  Basically it
=93attracts=94 the ambient field in a high permeability core, then satura=
tes
that core with an auxiliary winding =96 the driver winding -- causing its=

permeability to approach that of air, therefore the previous ambient
field lines in the core are =93pushed=94 out.  If the core is surrounded =
by
another coil which is magnetically orthogonal to the driving coil =96 the=

sense coil -- the sense winding is not coupled magnetically to the drive
winding, but the flux from the ambient field that is moving in and out
induces voltages in the sense coil.  A little thought about this process
shows that although the drive coil saturates on a positive and negative
drive peaks, the result on the sense coil is that the sensed voltage has
gone though two positive peaks, i.e., it is the second harmonic =96 it is=

two times the frequency -- of the drive waveform.  This is why flux gate
magnetometers are often called harmonic modulators.  Typically to
improve sensitivity, the sense winding is resonated with a capacitor to
the second harmonic frequency.

The Speake and Precision Navigation magnetometers use another effect.
Suppose we take a strip of high permeability metal that is easily
saturated (easy saturation is not actually necessary, it just makes the
effect easier to use.)  Now we wrap a coil around this strip of metal
and put it in an oscillator circuit that senses the BEMF in the coil.
What happens next is that the BEMF collapses when the strip saturates
under the drive current.  If this is the signal to reverse the current
in the drive coil, then the unbalance in the drive current is the
measure of the ambient flux.  The Speake and Precision Navigation
circuits do one more slight variation.  Instead of symmetrically sensing
and reversing the drive current, they reset the core in one direction
very quickly and only sense the saturation in the opposite direction.
The result of this tactic is that the time to saturation varies with the
ambient field and since this time is the pulse width of the output,
therefore the frequency varies roughly proportional to the ambient
field.

I=92ve been developing a high accuracy, low cost version of the
symmetrical form that can be read out with one of the 24 bit DACs that
are now so available with a very stable zero and symmetrical plus and
minus sense range.  The strip of permalloy or Metglas is free to $0.25.
A hint for gathering the strips: save all your anti-theft tags such as
on CD=92s from Best Buy and other retailers, and make good friends with
your local librarian.  A curious thing, I have picked up 20 or 30 the
tags in the parking lot of my local Best Buy store.  I can only surmise
that as people leave the store they instantly rip off the cellophane
wrapping, along with the tags, in order to start playing their new CD=92s=

on the way home.  I can easily pick-up 5 to 10 tags in a single trip to
the parking lot.  Just look down.  Current results on the librarian
source strips are a noise floor of about 50 pT peak-to-peak to about 30
Hz not counting ambient hum.  It=92s actually capable of about a 1 KHz BW=
,
but I close it down to keep the 60 HZ pick-up manageable.  Probably the
anti-theft tags, which use Metglas, will not perform as well as the
library permalloy strips since the Metglas in the tags is probably
annealed to be highly magnetostrictive.  This quality interferes with
the noise performance from what I=92ve read.  I=92ve haven=92t finished t=
he
experiments in this area to confirm that.

Regards,
Charles R. Patton



_____________________________________________________________________

Public Seismic Network Mailing List (PSN-L)